Prabir Sarkar
S. Phaneendra
Amaresh Chakrabarti
1
e-mail: ac123@cpdm.iisc.ernet.in
Innovation, Design Study and Sustainability
Laboratory (IDeaS Lab),
Centre for Product Design and Manufacturing,
Indian Institute of Science,
Bangalore 560012, Karnataka, India
1
Developing Engineering Products
Using Inspiration From Nature
Nature can be a major source of inspiration for engineering designers. Biomimicry is
often used in specific cases to develop solutions that mimic natural systems. However,
knowledge of natural systems is still not used systematically and commonly for inspiring
innovative product development, from ideation of solutions to their implementation as
products. In ideation, potential solutions to a design problem are generated. To support
ideation, two databases are developed with entries having information about natural and
artificial systems. A novel generic causal model is developed for structuring information
of how these systems achieve their behavior. Three algorithms are developed for analogical search of entries that could inspire ideation of solutions to a given problem. In
realization, evaluation and modification of these solutions are carried out by experimenting with these in virtual and physical forms and environments.
关DOI: 10.1115/1.2956995兴
Introduction
Designs found in nature are wonderful. Nature, through billions
of years of trial and error, has produced effective solutions to
innumerable complex real-world problems 关1兴, whose functions
are often similar to that demanded in engineering products, e.g.,
heart functioning like a pump 关2兴 or joints with movement as in a
mechanism 关3兴. For both natural and engineering tasks, resources
are limited and must be utilized optimally to accomplish the task
in a reliable and functional manner. Thus, designs in nature could
act as inspirations for the engineering designer 关4兴. The question
is, how to use systematically, the knowledge of these systems to
solve design problems?
Vincent 关1,5兴 estimates that “at present there is only a 10%
overlap between biology and technology in terms of the mechanisms used” so there is a great deal of potential in this area.
Engineers, scientists, and businesses are increasingly turning toward nature for design inspiration. The field of biomimetics, the
application of methods and systems found in nature to engineering
and technology, has spawned a number of innovations far superior
to what the human mind alone could have devised, such as a solar
cell inspired by a leaf 关6兴. Its focus has primarily been on longterm development of specific technologies such as synthetic spider
silk 关7兴. However, there is no general systematic support available
for engineering designers to use nature as inspiration for solving
design problems—right from ideation to realization of ideas.
We developed a two-step approach to address this issue. The
first step is “ideation,” where potential solutions to a problem are
generated. The second step is “realization,” where these are evaluated and modified by experimenting with them in both virtual and
physical forms.
2
Ideation
Inspiration is useful for exploration and discovery of new solution spaces 关8兴. Evidence of this, for instance, are that the presence of a stimulus can lead to generation of more ideas during
problem solving 关9兴, that stimulus-rich creativity techniques positively affect creativity 关10兴, or that individuals stimulated with
association lists demonstrate more creative productivity than those
without stimuli 关11兴.
1
Corresponding author.
Contributed by the Computer-Aided Product Development 共CAPD兲 Committee of
ASME for publication in the JOURNAL OF COMPUTING AND INFORMATION SCIENCE IN
ENGINEERING. Manuscript received September 27, 2006; final manuscript received
June 2, 2008; published online August 5, 2008. Assoc. Editor: R. Crawford.
Analogy has long been regarded as a powerful means for inspiring novel idea generation, as seen in several systems based on
analogy 关12,13兴 and creativity methods developed with the specific aim of fostering analogical reasoning 关14兴, and in other research that have contributed to the understanding and application
of analogy 关15–18兴.
Generating a large number of ideas with a variety of principles
关19兴 increases the chances of developing better products 关9兴; experimenting with ideas in realistic environments makes them more
practically viable. Creative products are initiated through inspiration 关3,12兴, where use of analogy is beneficial 关1,14,20兴.
Shu et al. 关21兴 and Shu 关22兴 shown the application of a biomimetic search method to develop ideas on how to center objects in
microassembly and remanufacturing 关23,20兴. Structural analogy
has been used to achieve this. Vakili and Shu 关23兴 worked on
biomimetics where analogous keywords of a system’s function are
used to help designers. Single text with synonyms has been used
for searching analogous functions. Later Hacco and Shu 关20兴 used
Wordnet® and text tags to help find similar words and part of
speech in a sentence, similar to natural language processing.
Based on the structure of the system, solutions are generated related to remanufacturing.
In our work we have used a combination of words for searching
entries and this search can be carried out with single or a combination of SAPPhIRE constructs 共see Sec. 2.3兲. At a broad level of
abstraction, our work is similar to Hacco and Shu 关20兴, as we both
use synonyms to find alternative solutions. However, we argue
that the approach presented here is far more extensive in three
ways. First is, unlike Vakili and Shu 关23兴, where only functional
analogy has been used, we use functional, behavioral, and structural analogies, using seven layers of constructs from the
SAPPhIRE model of causality to carry out the search. The second
is combination of analogies from multiple layers can be used in
our case to carry out search 共see Sec. 2.4兲. Finally, results of
search using multiple analogies can be used in the analogical reasoning itself 共see Sec. 2.4兲.
2.1 Methodology. Earlier, we developed a computational tool
called IDEA-INSPIRE that provides analogical ideas of natural or
artificial systems as inspirations to designers to support generation
of novel solutions for product design problems 关4兴. The current
application focus is novel mechanisms. Our intention is to use the
diverse motions that nature exhibits as a source of inspiration for
solving product design problems, especially in inspiring creativity
and innovation of novel products. The work is not about mimicking natural phenomena but rather getting inspired from primarily
the behavioral aspects of these phenomena.
Journal of Computing and Information Science in Engineering
Copyright © 2008 by ASME
SEPTEMBER 2008, Vol. 8 / 031001-1
Downloaded 27 May 2010 to 203.200.43.195. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
To develop IDEA-INSPIRE, first two databases were developed:
one for natural systems 共e.g., insects, plants, etc.兲 exhibiting diverse movements and the other for artificial systems 共e.g., vacuum
cleaners, clutches, etc.兲. These systems were then analyzed, and a
generic SAPPhIRE model of causality for describing their behaviors was developed. This model has a set of constructs that are
used to explain the functioning of an artificial or natural system.
Finally, IDEA-INSPIRE software was developed with appropriate interface and analogical search procedures to aid the following. Designers, with a problem to solve, would explore the behavior of
the natural or artificial system entries in the databases, and use the
constructs of the SAPPhIRE model to describe the problem in
terms of the constructs of the language; the software would then
search the databases for entries that could be presented to the
designers as inspirations to aid ideation to solve the problem.
IDEA-INSPIRE is based on the following philosophy: given a design
problem, if the designer is exposed to a variety of natural and
engineered systems that have similar function, behavior, or structure, ideation should be enhanced.
2.2 SAPPhIRE Model of Causality. The main challenge in
developing the model of causality was that it must allow the function, behavior, and structure of a system to be linked to one another in a way common for both natural and artificial systems, and
allow describing these at various levels of abstraction. At the center of this work is the development of a uniform functional/
behavioral representation for these systems.
We view function as the intended effect of a system 关24兴 and
behavior as the link between function and structure defined at
given levels; therefore, that is the behavior specific to the levels at
which the function and structure of a system are defined. We argue
that structure—in a richer representation of causality—must have
the flexibility of being represented using multiple views. Such a
richer encompassing causal description of the functioning of a
system has been developed and described below 关4兴.
The constructs used in the SAPPhIRE model of causality are as
follows.
Parts. Physical components and interfaces constituting the system and its environment of interaction.
State. Attributes and their values that define the properties of a
given system at a given period of time during its operation.
Physical effect. The laws of nature governing change.
Organ. The structural context necessary for a physical effect to
be activated.
Input. The energy, information, or material requirements for a
physical effect to be activated.
Physical phenomenon. Potential changes associated with a
given physical effect for a given organ and inputs;
Action. Abstract description/interpretation of a state change, a
changed state, or an input.
The relationships between these constructs are as follows: parts
are necessary for creating organs. Organs and inputs are necessary
for activation of physical effects. Activation of physical effects is
necessary for creating physical phenomena, which activates and
changes of state. Change of state is interpreted as actions or inputs, and create or activate parts. Essentially, there are three relationships: activation, creation, and interpretation 关4,25兴.
The model is acronymed as the SAPPhIRE model, where
SAPPhIRE stands for state-action-part-phenomenon-input-organ
effect 共see Fig. 1兲.
2.3 Representation. Each entry 共e.g., in the Appendix兲 in the
databases of IDEA-INSPIRE 关4兴 is described using entry-specific details that are structured using its function, structure, and behavior,
as well as using a linked set of SAPPhIRE constructs, videos and
images 关4兴. The databases 关4兴 have over 700 entries. IDEA-INSPIRE
is implemented using Microsoft™ VISUAL C⫹⫹.
Each entry in the databases is described, using the SAPPhIRE
constructs 共see Appendix兲, in two forms.
1. A computer understandable form—these are text files con031001-2 / Vol. 8, SEPTEMBER 2008
Fig. 1
The SAPPhIRE model of causality
taining a set of words used to describe the entry and formatted in a particular way that can be read by the program
easily and can be used for efficient search.
2. A human understandable form—these are text files containing paragraphs, explaining the motion behavior of the entry
in sentence form.
In the computer understandable form, the content of each entry
is divided into a list of actions, state, physical phenomena, inputs,
physical effects, organs, and parts; links between various subsets
of these together constitute the entry.
In the human understandable form of the entries, these links are
expanded in paragraphs 共see Appendix for an example兲.
2.4 Search Strategies. The goal of the search strategies is to
support analogical reasoning at multiple levels of abstraction. Using these strategies, a designer should be able to simply browse
the entries for random stimulation or systematically search
through them with specific purposes. For searching for solutions a
designer can use one or many constructs 共action, physical effect,
etc.兲 of the SAPPhIRE model, or any combinations of them as
input to the software. Presently the software can take three demands and three wishes. Each demand or wish can be depicted by
any of the seven constructs. Each of these constructs has a particular way of expressing—verb, noun, and adjective or phrases. A
designer can use many combinations of these constructs in order
to completely define the problem.
Three search algorithms are developed: simple, combination,
and complex searches.
Simple and combination search. While searching for analogical
entries, a designer can give one 共simple search兲, or a combination
of 共combination search兲 constructs 共action, physical principle,
etc.兲 to the software as inputs 共see Appendix for an example of an
entry兲.
Complex search. Since each entry is a linked network of SAPPhIRE constructs, analogical solutions could be reached if IDEAINSPIRE could be searched for, say, entries that share the same
principles in the entries that fulfill a given required action or entries that have analogical parts to those having the required action.
2.5 Working Principle. The description of characteristics or
constraints obeyed by entries looked for is provided by the designer in terms of verbs, nouns, and adjectives. In order to implement the “translation” of this input into analogical descriptions,
clusters of equivalent words 共synonyms兲 have been developed.
Clustering of words is carried out beforehand for nouns, verbs,
and adjectives and stored in a database for use during translation.
When a designer gives the required input to the software by
formulating the problem in terms of a combination of constructs,
the software tries to find the best match of the entries in the
Transactions of the ASME
Downloaded 27 May 2010 to 203.200.43.195. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Table 1 Efficacy of
Without using
software „D1–D3: designers involved…
With using
IDEA-INSPIRE
IDEA-INSPIRE
Ratio of
S2/G2
No. of
entries
explored
共E兲
Ratio of
G2/E
6
35%
60
28%
5
3
60%
40
12%
37%
18
6
33%
60
30%
49%
13.3
5
42%
53
23%
Ratio of
S1/G1
No. of
ideas
generated
共G2兲
No. of
ideas
selected
共S2兲
4
44%
17
6
4
66%
8
3
7.6
3.6
No. of
ideas
generated
共G1兲
No. of
ideas
selected
共S1兲
D1
9
D2
D3
Avg.
IDEA-INSPIRE
descending order of importance, annotating each entry with a
weightage as described in Sec. 2.4, which shows how close the
entry is to the search input provided.
Since each entry is a linked network of actions, state changes,
organs, inputs, effects, phenomena, and parts, analogical solutions
can be reached if it is possible to search for, say, entries that share
the same principles in the entries that fulfill the required actions or
entries that have analogical parts to those that have the required
intended action. For search of these kinds, one should be able to
construct complex search queries with multiple, intermediate, and
final search points of specified types and with specified input
types and values. This can be achieved by a multiple search with
the following form.
For a given input type, find all outputs of a given type for all
intermediate outputs of given types. For instance, one such query
is for a given action 共input兲, find all entries 共output兲 that provide
actions 共intermediate output type兲 that use the same effects 共intermediate output type兲 as are used by the entries that provide the
input action.
Given action→effects used→actions→entries. Such complex
search problems should help “discover” solutions that are more
difficult to immediately associate with a given design problem
共e.g., the input action兲 and yet are analogically relevant as potential ideas for solving the problem.
2.6 Evaluation of the Ideation Support. Ideation effectiveness of the support was evaluated in two stages. In the first stage,
three designers were asked to individually generate as many ideas
as possible to solve design problems without any aid. They were
then asked to generate further ideas taking inspiration from entries
Ratio
of
G1/G1
188%
共17/ 9兲
083%
共5 / 6兲
225%
共18/ 8兲
165%
in
IDEA-INSPIRE.
These designers were provided with an instruction sheet that
described the problem to be solved and were requested to sketch
and annotate the ideas generated by them, identifying the major
components for each idea and explaining its working principle.
Each designer was provided with a set of numbered blank sheets
and was requested to sketch each idea on a different sheet. First,
each solved all three problems without aid. Next, each evaluated
their ideas, and selected some of them as acceptable solutions.
IDEA-INSPIRE software was then introduced to each designer by
the researcher. For each designer and problem, a number of search
strings were formulated using input from the designer involved,
and IDEA-INSPIRE was used for search of entries to inspire ideation
for each problem. The entries were seen by the designer one after
another, and if the designer get inspired, 共s兲he created an idea
immediately based on the entry seen. This continued for several
search sessions for each problem until the designer expressed termination of this process. There was no time constraint prespecified for any session. After the ideation sessions for a given problem for each designer, the resulting ideas created by the designer
were evaluated by the designer, and acceptable solutions among
these were selected. Since each idea was presented on a separate
numbered sheet, counting the number of ideas did not pose a
problem. In any case, the ideas were counted by two individual
researchers for accuracy. Each individual ideation session took
about an hour, and evaluation/selection about half an hour, taking
about 3 h altogether for each designer per problem solving, including ideation unaided, evaluation and selection of these ideas,
ideation aided by IDEA-INSPIRE, followed by evaluation and selec-
Fig. 2 Some of the initial concepts that the designers generated while solving Problems 1 „left… 2 „right…
Journal of Computing and Information Science in Engineering
SEPTEMBER 2008, Vol. 8 / 031001-3
Downloaded 27 May 2010 to 203.200.43.195. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Fig. 3 An entry of a bee as the source of inspiration „left…. Simulation of the intended vehicle
moving forward, colliding with the obstacle, retracting after collision, and taking a left turn to
avoid the obstacle „top right…. Physical modeling of how it avoids an obstacle „bottom right….
tion of these ideas.
On average, each designer was able to generate 165% additional ideas using inspiration from IDEA-INSPIRE than on their own
共see Table 1兲, which indicated the ideation effectiveness of the
approach. This was assessed using two parameters: one is the
enhanced fluency of the designers, as seen from the additional
ideas generated by them. This is a standard measure used by many
creativity researchers in assessing likely novelty of the outcomes,
such as in Refs. 关26–28兴. The second is the percentage of ideas
selected by the designers themselves as worth developing 共about
42%兲, which indicates the likely usefulness of the ideas generated
关28,29兴. As described by the majority of creativity researchers
关26–29兴, and summarized in our earlier work on developing a
common definition of creativity 关30兴, novelty and usefulness are
the two basic components of the creative quality of ideas. Therefore assessing ideation effectiveness using these parameters implies the impact of the software in enhancing the creative quality
of ideas. This, we argue, happens due to exploration of a larger
number of entries by the designers, which get them triggered with
newer ideas with varied characteristics, helping them generate
greater variety.
In the second stage of evaluation, 12 design students who were
in their final stage of their course were offered support of ideation
using IDEA-INSPIRE in their respective design problems that they
had to solve as a mandatory requirement for their course. The
designers participating in these evaluation stages have an undergraduate degree in engineering, up to a year of professional experience in design in industry, and are in the final year of a two year
Masters in Design course at the time of the evaluations. Feedback
from this second informal evaluation 共based on their exit comments on the usefulness and usability of the software for their
purpose兲 indicated both an enhancement of the number of new
ideas generated 共as in the first evaluation兲 and ease of use of the
software.
This, however, does not support attainment of realizability of
these ideas—an essential aspect of engineering product development. For this, a realization support was developed and used in
another study, where working prototypes to embody these ideas
were created.
Fig. 4 Frogs hind legs and Bat’s wings as sources of inspiration for a foldable mechanism
„top left…; simulation of deployment folding mechanism „top right…; physical model and sequence of operations „bottom…
031001-4 / Vol. 8, SEPTEMBER 2008
Transactions of the ASME
Downloaded 27 May 2010 to 203.200.43.195. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Fig. 5 Japanese fan as a source of inspiration „left…; simulation and physical model of deployment
„right…
3
Realization
The realization support has the following parts.
1. Virtual modeling and experimentation support, which uses
SOLID WORKS™, MSC VISUAL NASTRAN™, and MATLAB’s simulink™ to support simulation and experimentation of the
essential subsystems in a variety of virtual environments.
Each design underwent several iterations and modifications before being selected for physical modeling and experimentation.
2. Physical modeling and experimentation support uses LEGO
MINDSTORMS™ robotic kit. A scaled model of the required
system with the essential subsystems is modeled. In physical
modeling, input values for various parameters were taken
from the virtual simulations in order to test physical prototypes for functioning.
3.1 Modeling of the Entire System: A Step Toward Actual
Working Prototype. The realization support helps fast experimentation with ideas using, respectively, a virtual and a physical
modeling environment. Using virtual modeling first allows experimentation with a wider set of ideas faster, thereby pruning the
number to a few that pass the test, while physical modeling,
though time consuming, enables a justification better grounded in
reality.
These experiments are used to evaluate the usefulness of the
realization support. This is done in terms of how well the support
helps representation of the task and the environment, aids creation
of realistic solutions and their simulation in realistic environments, and the resulting inspiration it provides to designers to
make the ideas more realizable.
3.2 Evaluation of Realization Support. We demonstrate the
evaluation process with a case study, in which two experienced
designers were given two problems to solve individually using the
two-step approach and the support developed. Both designers
have formal graduate training in designing and have two to three
years of experience in designing engineering products in industry.
Neither had any prior experience in solving the kind of problems
used in the experiments described below.
In both the cases, the designers first used IDEA-INSPIRE to generate a variety of ideas for each problem, and selected some of
these for further development. Then, they used the virtual modeling support to model these ideas as solutions and used virtual
simulation to evaluate, modify, and select some for further development. These solutions were then modeled in the physical modeling environment for evaluation, modification, and final selection. The result is creation of product ideas that are both novel and
realistic.
The problems used in the experiments were to
1. design and develop an all terrain vehicle that can travel in
air, on the ground and under water. The vehicle needs to take
appropriate routes based on the presence and size of an obstacle and
2. design and develop a remotely controlled deployable mechanism for a solar panel for use in space. The area of the solar
panel should be as large as possible and the entire system
should be foldable in the least amount of space.
By inspecting the entries retrieved 共e.g., bee, rockets, etc.兲 from
each designer was able to generate a variety of initial solutions 共see Fig. 2兲. Each problem was expressed using various verb-noun-adjective combinations, e.g., move共V兲 + solid共N兲
+ medium共A兲 or none共V兲 + motion共N兲 + none共A兲. The software retrieved many entries, and the designers generated various ideas for
solving the problem by taking various aspects of these entries. The
IDEA-INSPIRE,
Table 2 Statistics on the problems solved „D1, D2: designers; VS/PS: virtual/physical
simulation…
No. of
solutions
generated
during
ideation
Time
taken in
hours for
ideation
共h兲
No. of
solutions
chosen for
VS
D1
D2
5
6
2
2.5
1
2
D1
D2
11
5
3
1.5
2
1
Time
taken
for VS
共h兲
No. of
solutions
chosen for
PS
No. of
iterations
during PS
Total
amount
of time
for PS
共h兲
Problem 1
4
3
15
12
1
1
3
3
3
3
Problem 2
3
2
10
4
2
1
3
1
4
2
No. of
iterations
during VS
Journal of Computing and Information Science in Engineering
SEPTEMBER 2008, Vol. 8 / 031001-5
Downloaded 27 May 2010 to 203.200.43.195. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Fig. 6 Some of the initial intended systems
entries retrieved by the software were from both artificial and
natural domains. From the natural domain, some of the ideas retrieved were bee, butterfly, housefly, and locust. From the artificial
domain, some ideas retrieved were rocket, steam engine, disk
brakes, and cam mechanism. One such entry is shown in the Appendix. Next, designers selected some of these ideas, modeled
them, simulated them in different environmental conditions, and
modified, leading to their progressive pruning to those which in
modified form fulfilled the requirements of the design to their
satisfaction.
3.3 Virtual Modeling and Simulation. The first solution is
an all-terrain, all-medium vehicle, with four wheels to maneuver
on land, an inflatable balloon to use as ballast to go down under
water, and a set of deployable legs to enable walking avoiding
small obstacles on land or under water.
Virtual modeling and simulation were carried out in steps. Motion on land and obstacle avoidance were part of the simulation.
Pressure sensors were attached to different parts of the vehicle,
aiding obstacle avoidance. A motor was attached to each wheel
and the constraints between the wheels of the vehicle and the
ground plane were specified. Rotation of the wheels was controlled by using MATLAB’s simulink™. In another version, the vehicle had a proximity sensor and avoided obstacles without colliding on them 共see Fig. 3兲. Inputs were taken from IDEA-INSPIRE
for modification, as required.
The designers created two more solutions which were also virtually modeled, simulated, and made into physical prototypes; the
first one used inspiration from a millipede and for the second one
from a leech.
3.4 Physical Modeling and Simulation. Dimensions of the
components and values of the variables 共e.g., frictional force required, torque at the wheels, placement of sensors, etc.兲 were finalized during virtual simulation. Taking these values, physical
modeling of the designs was carried out using LEGO MINDSTORMS™ building blocks. The modeled designs were controlled
by RCX, a programmable device, which is part of the LEGO MINDSTORMS™ kit.
For the second problem, the designers followed a similar sequence of steps as in the first problem. Using IDEA-INSPIRE software, they generated initial solutions and selected some of these,
Fig. 7 An all terrain vehicle
Fig. 8 Use of different systems of an all terrain vehicle
031001-6 / Vol. 8, SEPTEMBER 2008
Transactions of the ASME
Downloaded 27 May 2010 to 203.200.43.195. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
which they modeled using the modeling support. Then, simulation
was carried out using inputs from IDEA-INSPIRE and from designers
themselves. The final selected solutions were modeled, simulated,
and developed into physical prototypes 共see Figs. 4 and 5 for the
source of inspiration, simulations and physical prototypes, and the
intended resulting design for two such solutions兲.
Table 2 shows the number of solutions generated in the problems solved, and the amount of time spent in ideation, visualization, and evaluation of these solutions for various stages of development of the ideas. Designers were able to generate, explore, and
evaluate some 27 initial solutions in a span of 9 h 共about 20 min
per idea兲. They could model, explore, and modify the six solutions
selected, on the virtual environment, in 41 h 共about 6.83 h per
solution兲, and finally model, explore and modify the six solutions
selected in a span of 12 h 共average 2.4 h兲. Together five realistic
solutions were developed during a course of 62 h 共just under 8
working days兲. On an average it took 3.4 h for each virtual iteration and 1.2 h for each physical iteration. From our own experience of developing industrial products, we believe that this could
be considered as a fast ideation-realization process.
Note that some of these designs already exist 共like the fan
model for deployment of solar panels, which was designed for
Phoenix Mars Lander 关31兴兲. However, the fact that the designers,
with no experience or interest in this domain of design, ignorant to
prior existence of these designs, were still able to generate them
independently is a testimony to the support of creativity provided
by the methodology. This personal creativity or p-creativity as
described by Boden 关32兴, which means the ability to create an idea
that is novel to the individual generating it, is an essential prerequisite to the eventual desirable s-creativity or societal creativity
关32兴 where ideas are novel to the entire society.
Apart from these design experiments, many other designers
have used the software to generate solutions for their own design
problems 共which were not physically modeled兲, such as design of
a space station repair vehicle 共used inspiration from leech that can
stick to the wall or hang and move even in the absence of gravity兲,
control of high temperature, high pressure fluid flow 共used inspiration from the various valves found in animal bodies and in water
supply circuits兲, and cutting mechanisms 共used inspiration from
various cutters used in machine shops as well as from barnacles
and sunflower兲.
The two-step approach described in this paper supports both
“generative variation” 共through ideation兲 and “adaptive variation”
共through realization兲, as described by Fricke 关33兴 as two major
strategies of designers for solution generation. Generative variation has been found by earlier researchers to enhance novelty of
solutions by enabling exploration of a large space of ideas, while
adaptive variation enables their consolidation and optimization,
making them more implementable and realistic, as emphasized by
Thompson and Lordan 关34兴. The overall effect is development of
ideas with both novelty and realizability. Some of the initial designs for Problem 2 共Sec. 3.2兲 are in Fig. 6.
3.5 Learning From Modeling and Simulation. A number of
instances of learning took place during use of the modeling support. In most of the cases, designers used the details provided in
the relevant entries from IDEA-INSPIRE. Some of the things designers learned about during simulation are as follows.
Friction and restitution coefficients necessary for a car to move
freely on the ground.
The linear momentum with which the car should collide with
the obstacle so that the touch sensor would be actuated but the car
would not undergo serious deformation, and the extent to which
the wheels of a car should rotate for it to avoid obstacles.
Legs of a vehicle need a mechanism similar to feet, which must
be dynamically balanced, else it could not move forward or retract. This led to exploration of several ways by which the body of
the vehicle can be dynamically balanced.
For this mechanism to perform appropriately, the motion of the
legs must follow an appropriate sequence.
Fig. 9 Solar array deployment „left… and deployed condition in
intended situation „right…
3.6 Modeling of the System. It can be noticed from Figs. 4
and 5 that even though the systems are realized, the structural
components still do not represent the actual system. So, the selected solutions, including their major subsystems, are further
modeled using CAD software. These models are explained next.
The solution for Problem 1 共Fig. 7兲 is an all terrain vehicle. Figure
7 shows its major components 共see Fig. 8 for the simulated
model兲.
The major components are a balloon for flying in air and two
cylindrical chambers that can be filled with water to control the
vehicle to be at various heights in water. A set of deployable
wheels enables it to run on ground. The deployable walking legs
help it to climb over big obstacles and walk on river bed, clinging
to something when the water current is high. Figure 8 shows the
use of its major subsystems in different situations.
Two solutions for Problem 2 共Figs. 9 and 10兲, earlier modeled
in Figs. 4 and 5, are two solar array deployment systems with the
main structural components modeled.
Modeling the entire system gives a framework for further design and more detailed modeling by domain experts. Physical
modeling provides a more concrete form to each solution. The
final system, being complex, cannot be tested before an actual
prototype is developed, requiring many hours of subsystem modeling by experienced engineers.
In an earlier work done by Sarkar and Chakrabarti 关35兴, various
novelty measuring methods have been reviewed, and a new
method for measuring creativity has been proposed and evaluated.
This method uses novelty and usefulness of a product to determine creativity of the product. Using this method one can categorize the novelty of products into four groups: very highly novelty
共its function itself is novel兲, high novelty 共input and state changes
used by the product are novel兲, medium novelty 共physical phenomena or effects used are novel兲, and low novelty 共organs or
parts used are novel兲. We have found that the ideas generated in
the cases studied in this paper are in high and medium novelties.
4
Conclusions
We hope to have demonstrated that systematic use of knowledge from both artificial and natural domains can not only help
designers to generate a variety of solutions but also aid in their
development into realizable and practical prototypes. Using experimental studies of engineering designers we explained the
workings of a biomimetics-based framework for systematic support for designers to develop novel realizable engineering solu-
Fig. 10 Solar array deployment „left… and deployed condition
in intended situation „right…
Journal of Computing and Information Science in Engineering
SEPTEMBER 2008, Vol. 8 / 031001-7
Downloaded 27 May 2010 to 203.200.43.195. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Fig. 11
An entry from the database
tions using inspiration from natural and artificial domains. The
methodology and support developed should help engineering designers to use inspiration from nature in order to generate products
as solutions to engineering problems and then develop them into
practical prototypes. Indian Space Research Organization 共ISRO兲
has delivered IDEA-INSPIRE to help its designers to generate suitable products for space applications. A customized version of
IDEA-INSPIRE is also delivered to the innovation wing of a global
engineering conglomerate to inspire their designers in ideation.
SAPPhIRE Constructs
•
•
•
Acknowledgment
The authors acknowledge the help rendered by the designers
who participated in the evaluation of the methodology. Microsoft™ is registered to Microsoft Corporation. SOLID WORKS™
is a registered to SolidWorks Corporation. MSC VISUAL NASTRAN™
is registered to MSC.Software Corp. LEGO MINDSTORMS™ is registered to Lego group. MATLAB’s simulink™ is registered to Mathworks, Inc. ISRO supported earlier development of this methodology. All authors contributed equally to this work and declare no
competing financial interest.
Appendix: Example Showing Human Understandable
(FBS and SAPPhIRE Model) Codes for an Entry (Bee)
in the Database
FBS
Function. Bees fly in order to transport themselves from place
to place in search of food. They mainly consume nectar of the
flowers.
Structure. Bees have two muscle systems to enable flight: direct
muscle system and indirect muscle system. Figure 11 shows the
double-hinged attachment of the wings to the thorax, one hinge
being connected to the side of the thorax 共outer hinge兲 and the
other to the tergum 共inner hinge兲. The dorsoventral muscles run
from the tergum to the bottom of the thorax.
Behaviour. The bees employ the indirect muscle system. The
dorsoventral muscles, running from the tergum to the bottom of
the thorax, contract to raise the wings 共starting from the image at
the bottom兲. When the dorsoventral muscles contract, the tergum
is lowered and the wings rotate about the outer hinges and rise.
The longitudinal muscles, running along the length of the thorax,
contract to lower the wings. When the longitudinal muscles contract, the tergum is forced upward again, and the wings rotate in
the opposite sense about the outer hinges. The effect of the indirect musculature may be described by a familiar object: the tennis
ball.
031001-8 / Vol. 8, SEPTEMBER 2008
•
•
•
•
Action
To fly from one place to another.
To move from one place to another.
State
The insect is at rest.
The insect is flying.
Phyphenomenon
The dorsoventral muscles contract.
The wing rotates in the clockwise direction.
The longitudinal muscle contracts.
The wing rotates in the counterclockwise.
The wing movements are monitored by the brain.
Phyeffect
Hooks law.
Lever effect.
Input
The contracting force.
The moment applied.
The electrical signals.
Organ
Link between the wing and the body.
Link between the wing and the muscle.
Link between brain and body.
Parts
Dorsoventral muscles.
For an example of Machine-understandable codes for an entry, see
关4兴.
References
关1兴 MONGABAY™, http://news.mongabay.com/2005/0711-rhett-butler.html.
关2兴 Vogel, S., 1992, Vital Circuits, Oxford University Press, Oxford.
关3兴 Vogel, S., 1998, Cat’s Paws and Catapults: Mechanical Worlds of Nature and
People, Norton, New York.
关4兴 Chakrabarti, A., Sarkar, P., Leelavathamma, and Nataraju, B. S., 2005, “A
Functional Representation for Biomimetic and Artificial Inspiration of New
Ideas,” Artif. Intell. Eng. Des. Anal. Manuf., 19, pp. 113–132.
关5兴 BIONICS™, http://en.wikipedia.org/wiki/bionics.
关6兴 Benyus, J. M., 1997, Biomimicry: Innovation inspired by nature, Quill.
关7兴 David, G., 1994, “Synthetic Spider Silk,” Technol. Rev., 97共7兲, pp. 16–17.
关8兴 Murakami, T., and Nakajima, N., 1997, “Mechanism Concept Retrieval Using
Configuration Space,” Res. Eng. Des., 9, pp. 9–11.
关9兴 Kletke, M. G., Mackay, J. M., Barr, S. H., and Jones, B., 2001, “Creativity in
the Organization: The Role of Individual Creative Problem Solving and Computer Support,” Int. J. Hum.-Comput. Stud., 55, pp. 217–237.
关10兴 MacCrimmon, K. R., and Wagner, C., 1994, “Stimulating Ideas Through Creative Softwares,” Manage. Sci., 40, pp. 1514–1532.
关11兴 Watson, D. L., 1989, “Enhancing Creative Productivity With the Fisher Associated Lists,” J. Creat. Behav., 23, pp. 51–58.
关12兴 Bhatta, S., Goel, A., and Prabhakar, S., 1994, “Innovation in Analogical Design: A Model, Based Approach,” Proceedings of the AI in Design, Kluwer
Transactions of the ASME
Downloaded 27 May 2010 to 203.200.43.195. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Academic, Dordrecht, The Netherlands, pp. 57–74.
关13兴 Qian, L., and Gero, J. S., 1996, “Function-Behavior-Structure Paths and Their
Role in Analogy-Based Design,” Artif. Intell. Eng. Des. Anal. Manuf., 10共4兲,
pp. 289–312.
关14兴 Gordon, W. J. J., 1961, Synthesis: The Development of Creative Capacity,
Harper & Row, New York.
关15兴 Gentner, D., 1983, “Structure Mapping—A Theoretical Framework,” Cogn.
Sci., 7, pp. 155—177.
关16兴 Chiu, M., 2003, “Design Moves in Situated Design With Case-Based Reasoning,” Des. Stud., 24, pp. 1–25.
关17兴 Gentner, D., and Markman, A. B., 1997, “Structure Mapping in Analogy and
Similarity,” Am. Mineral., 52, pp. 45–56.
关18兴 Blanchette, I., and Dunbar, K., 2001, “Analogy Use in Naturalistic Settings:
The Influence of Audience, Emotion, and Goals,” Mem. Cognit., 29共5兲, pp.
730–735.
关19兴 Larry, W., Zimmerman, P. E., and Hart, D. G., 1988, Value Engineering: A
Practical Approach for Owners, Designers and Contractors, CBS, New Delhi.
关20兴 Hacco, E., and Shu, L., 2002, “Biomimetic Concept Generation Applied to
Design for Remanufacture,” Proceedings of 2002 ASME Design Engineering
Technical Conferences and Computers and Information in Engineering Conference, Montreal, Canada, Sept. 29–Oct. 2, Paper No. DETC 2002/DFM34177.
关21兴 Shu, L. H., Lenau, T. A., Hansen, H. N., and Alting, L., 2003, “Biomimetics
Applied to Centering in Microassembly,” CIRP Ann., 52共1兲, pp. 101–104.
关22兴 Shu, L., 2004, “Generalizing the Biomimetic Design Process,” Canadian Design Engineering Network Design Conference, Montreal, Quebec, July 29–30.
关23兴 Vakili, V., and Shu, L., 2001, “Towards Biomimetic Concept Generation,”
Proceedings of 2001 ASME Design Technical Conferences, Design Theory and
Methodology Conference, Pittsburgh, PA, Sept. 9–12, Paper No. DETC2001/
DTM-21715.
关24兴 Chakrabarti, A., and Bligh, T. P., 2001, “A Scheme for Functional Reasoning
in Conceptual Design,” Des. Stud., 22共6兲, pp. 493–517.
关25兴 French, M. J., 1998, Invention and Evolution: Design in Nature and Engineering, CUP, Cambridge.
关26兴 Sefertzi, E., 2000, Report Produced for the EC Funded Project Innoregio:
“Dissemination of Innovation and Knowledge Management Techniques.”
关27兴 Csikszentmihalyi, M., 1997, Creativity: Flow and the Psychology of Discovery
and Invention, Harper Perennial.
关28兴 Sternberg, R. J., and Lubart, I. T., 1999, “The Concept of Creativity: Prospects
and Paradigms.” in Handbook of Creativity, R. J. Sternberg, ed., Cambridge
University Press, Cambridge.
关29兴 McClelland, D. C., 1988, Human Motivation, 3rd ed., Cambridge University
Press, Cambridge.
关30兴 Sarkar, P., and Chakrabarti, A., 2008, “Studying Engineering Design Creativity,” Invited Paper in the Proceedings of the NSF Workshop on Studying Design Creativity, J. Gero, ed., Aix-en-Provence, France, March.
关31兴 FOURTH-MILLENNIUM™,
http://www.fourth-millennium.net/spaceexploration/space-exploration-phoenix.html.
关32兴 Boden, M. A., 1999, “Computer Models of Creativity” in Handbook of Creativity, R. J. Sternberg, ed., Cambridge University Press, Cambridge.
关33兴 Fricke, G., 1992, Proceedings Research in Design Thinking, N. Cross, K.
Dorst, and N. F. M. Roozenburg, eds., Delft University Press, Delft, pp. 105–
109.
关34兴 Thompson, G., and Lordan, M. M., 2001, “Experience of Creativity Enhancement Teaching in Industry and University” International Conference on Engineering Design ICED’01, Glasgow.
关35兴 Sarkar, P., and Chakrabarti, A., 2007, “Development of a Method for Assessing Design Creativity,” International Conference on Engineering Design,
ICED 2007.
Journal of Computing and Information Science in Engineering
SEPTEMBER 2008, Vol. 8 / 031001-9
Downloaded 27 May 2010 to 203.200.43.195. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm